One of the reasons I love d3 is that it's built on standards such as CSS and SVG/HTML. They are so widespread you can't go wrong learning them, and that knowledge is immediately usable elsewhere (other related benefits include extremely powerful graphic capabilities and tons of documentation).

For graphs beyond ggplot2, I am happy for the various D3 to R packages, such as

networkD3 (for Sankey diagrams) - https://christophergandrud.github.io/networkD3/ ,

R2D3 - https://rstudio.github.io/r2d3/ , and

htmlwidgets - http://www.htmlwidgets.org/